10 — 部署方案Deployment Plan
從單隊到鄉鎮規模。
From one team to township scale.
以下均為設計情境與成本試算，用於比較不同部署規模，不代表已完成相同規模的正式演練或採購。
These design scenarios and cost estimates compare deployment scales; they do not claim a completed drill or purchase at the same scale.
DESIGN SCENARIO / COST ESTIMATE
01
單隊演練情境
Single-team scenario
- →4 名搜救員、10 個節點、1 個 HQ 指揮端
- →規劃範圍 1 km²，任務時間 90 分鐘
- →同時處理 1 個 Victim SOS 與 1 個搜救員 MAYDAY
- →4 rescuers, 10 nodes and 1 HQ command endpoint
- →1 km² planned area and a 90-minute mission
- →One Victim SOS and one rescuer MAYDAY handled concurrently
02
設計估算DESIGN ESTIMATE
規模化與成本
Scaling & cost
- →10 節點：約 NT$20,000
- →40 節點：約 NT$80,000
- →100 節點：約 NT$200,000；均以 SBN 基準 NT$2,000／節點估算，不含人力、載具、SEN 價差與備品
- →10 nodes: approximately NT$20,000
- →40 nodes: approximately NT$80,000
- →100 nodes: approximately NT$200,000; calculated from the SBN baseline of NT$2,000 per node, excluding labour, vehicles, SEN cost differences and spares
03
鄉鎮級情境
Township-scale scenario
- →4 支搜救隊、40 個節點、2 個 HQ 端點
- →規劃範圍 4 km²，節點平均間距約 300–500 m
- →40 節點高負載預估抵達率 93.5%，P95 延遲 2.7 s
- →保留 10% 備援節點，即另備 4 個節點
- →4 rescue teams, 40 nodes and 2 HQ endpoints
- →4 km² planned area with 300–500 m average node spacing
- →40-node high-load estimate: 93.5% delivery and 2.7 s P95 latency
- →10% node reserve, equivalent to 4 spare nodes
後續優化方向Next steps
實驗驗證與硬體整合。
Field validation and hardware integration.
LinkGuard-E 已完成原型系統建置與初步功能驗證。若要進入實際救災，仍需補足通訊測試資料與硬體穩定性；以下為作品說明書中規劃的三個方向。
LinkGuard-E has a working prototype and initial functional validation. Real deployment still needs field link data and more robust hardware; these are the three directions planned in our report.
01
實際 RC 結構穿透試驗
Real RC penetration trials
- →單面 RC 試體：在學校土木科實習工廠的 RC 試體牆（約 15–20 cm）量測單層純 RC 的實際衰減，對照 Motley–Keenan 模型的每層 22 dB
- →多樓層穿透：接洽即將拆除的 3 層樓以上 RC 透天案場（避開地下停車場的 eTag／RFID 920–928 MHz 干擾），信標置於底層，接收端依序放上 1、2、3 樓；備案為戶外堆疊水泥磚加鋼筋網 1–3 層
- →模擬廢墟瓦礫：向新竹縣消防局申請借用倒塌建築模擬訓練設施，將信標埋入 0.5–2.0 m 不同深度，並驗證 RSSI 三角定位流程
- →校正：以實測建立「衰減量 → 樓層數」經驗對應表，取代固定 22 dB，重新代入蒙地卡羅模擬檢視 91.54% 是否需下修
- →法規：全程使用 NCC 合規 AS923 頻段，發射功率控制在 0.5 W 以下，測試前取得場域所有人書面同意
- →Single RC specimen: measure one-layer loss through a 15–20 cm RC test wall at the school's civil-engineering workshop and compare with the model's 22 dB per floor
- →Multi-floor: use an RC townhouse of three or more floors awaiting demolition (avoiding car parks, where eTag/RFID at 920–928 MHz interferes); beacon on the ground floor, receiver on floors 1–3; fallback is stacked concrete block with rebar mesh outdoors
- →Simulated rubble: borrow the Hsinchu County Fire Bureau's collapsed-building training facility, bury beacons at 0.5–2.0 m and validate the RSSI triangulation workflow
- →Calibration: build an empirical loss-to-floors table to replace the fixed 22 dB and re-run the Monte Carlo model to see whether 91.54% must be revised down
- →Compliance: NCC-compliant AS923 band throughout, transmit power kept under 0.5 W, written site-owner consent before testing
02
硬體整合：單一 PCB
Hardware integration: one PCB
- →現階段以 ESP32、LoRa 模組、電源模組、感測器與顯示元件跳線堆疊，接線過多、接點易鬆脫、外殼設計受限
- →設計整合 PCB：主控、LoRa 通訊、電源管理、感測器介面、狀態指示燈與充電保護整合至單一電路板
- →特別注意 LoRa 天線位置、射頻走線、接地配置、電磁干擾與電源穩定度
- →加入充放電保護、低電量偵測與電量回報，使信標與手持終端支援長時間任務
- →The prototype stacks ESP32, LoRa, power, sensors and display with jumper wires — too many connections, loose contacts and enclosure constraints
- →Design a single PCB integrating MCU, LoRa radio, power management, sensor interfaces, status LEDs and charge protection
- →Pay particular attention to antenna placement, RF routing, grounding, EMI and supply stability
- →Add charge/discharge protection, low-battery detection and battery reporting for long missions
03
NFC 傷票測試
NFC triage tags
- →每張 NFC 傷票具唯一識別碼，可讀取傷患編號、檢傷等級、生命徵象、處置紀錄與後送狀態
- →針對低光源、戴手套與多名救援人員交接的情境，測試連續掃描、重複讀取、資料寫入、離線暫存與同步
- →降低紙本傷票遺失、受潮或重複抄寫造成的問題
- →Each NFC tag carries a unique ID with patient number, triage level, vital signs, treatment record and evacuation status
- →Test continuous scanning, repeated reads, writes, offline caching and sync under low light, gloved hands and hand-offs between rescuers
- →Reduce the losses, water damage and re-copying errors of paper triage tags
完成此階段後，LinkGuard-E 將具備更完整的通訊測試資料與更穩定的硬體版本，為製作多套測試機、導入實際救災應用奠定基礎。
After this stage LinkGuard-E will have fuller link-test data and a more stable hardware revision, the basis for building multiple test units and moving toward real rescue use.